Elaboration of a method for comparison of Business Intelligence Systems which support data mining process

• Main Authors: Polańska, Julia, Zyznarski, Michał
• Format: Others
• Language: English
• Published: Blekinge Tekniska Högskola, Sektionen för datavetenskap och kommunikation 2009
• Subjects: Business Intelligence Systems; data mining; quality model; evaluation process; Software Engineering; Programvaruteknik
• Online Access: http://urn.kb.se/resolve?urn=urn:nbn:se:bth-2078

Business Intelligence Systems are becoming more and more popular in recent years. It is caused by the need of reusing data in order to gain some potentially useful business information about. Those systems are advanced set of tools, which causes high prices of purchase and licensing. Therefore, it is important to choose the system which fits the best particular business needs. The aim of this thesis is to elaborate a method for comparison of existing Business Intelligence Systems that are supporting data mining. The method consist of a quality model, build according to existing standards, and set of steps which should be taken to choose a Business Intelligence System according to particular requirements of its future user. The first part of the thesis focuses on the analysis of existing works providing a way for comparison of those software products. It is shown here that there is no existing systematic approach resolving this problem. However, criteria presented in those works along with the description of quality model standards were used for creating the quality model and proposing a set of basic measures. Also the phrases for the evaluation process were identified. The next part of the research is a case study which purpose is to show the usefulness of proposed evaluation method. The example is simple, but has proven that the method can be easily modified for specific needs and used for comparison of real Business Intelligence Systems. The quality level measured in the case study turned out to be very similar for each system. The evaluation method may be extended in future work with more advanced measures or additional characteristic which were not taken into account in this research.